bioc 2.0.dev2

BioC data structures and encoder/decoder for Python

BioC XML / JSON format can be used to share text documents and annotations.

bioc exposes an API familiar to users of the standard library marshal and pickle modules.

Development of bioc happens on GitHub: https://github.com/bionlplab/bioc

Getting started

Installing bioc

$ pip install bioc

XML

Encoding the BioC collection object collection:

import bioc

# Serialize ``collection`` to a BioC formatted ``str``.
bioc.dumps(collection)

# Serialize ``collection`` as a BioC formatted stream to ``fp``.
with open(filename, 'w') as fp
    bioc.dump(collection, fp)

Compact encoding:

import bioc
bioc.dumps(collection, pretty_print=False)

Incremental BioC serialisation:

from bioc import biocxml
with biocxml.iterwrite(filename) as writer:
    writer.write_collection_info(collection)
    for document in collection.documents:
       writer.write_document(document)

Decoding the BioC XML file:

import bioc

# Deserialize ``s`` to a BioC collection object.
collection = bioc.loads(s)

# Deserialize ``fp`` to a BioC collection object.
with open(filename, 'r') as fp:
    collection = bioc.load(fp)

Incrementally decoding the BioC XML file:

from bioc import biocxml

# read from a file
with biocxml.iterparse(filename) as reader:
    collection_info = reader.get_collection_info()
    for document in reader:
        # process document
        ...

# read from a ByteIO
with biocxml.iterparse(open(filename, 'rb')) as reader:
    collection_info = reader.get_collection_info()
    for document in reader:
        # process document
        ...

get_collection_info can be called after the with statement.

Together with Python coroutines, this can be used to generate BioC XML in an asynchronous, non-blocking fashion.

from bioc import biocxml

with biocxml.iterparse(source) as reader, biocxml.iterwrite(dest) as writer:
    collection_info = reader.get_collection_info()
    writer.write_collection_info(collection_info)
    for document in reader:
        # modify the document
        ...
        writer.write_document(document)

Json

Encoding the BioC collection object collection:

import bioc

# Serialize ``collection`` to a BioC Json formatted ``str``.
bioc.dumps(collection, BioCFileType.BIOC_JSON, indent=2)

# Serialize ``collection`` as a BioC Json formatted stream to ``fp``.
with open(filename, 'w') as fp
    bioc.dump(collection, BioCFileType.BIOC_JSON, fp, indent=2)

Compact encoding:

import bioc
bioc.dumps(collection, BioCFileType.BIOC_JSON)

Decoding the BioC Json file:

import bioc

# Deserialize ``s`` to a BioC collection object.
collection = bioc.loads(s, BioCFileType.BIOC_JSON)

# Deserialize ``fp`` to a BioC collection object.
with open(filename, 'r') as fp:
    collection = bioc.load(fp, BioCFileType.BIOC_JSON)

Json Lines

Incrementally encoding the BioC structure:

from bioc import BioCJsonIterWriter
with open(filename, 'w', encoding='utf8') as fp:
    writer = BioCJsonIterWriter(fp, level=bioc.PASSAGE)
    for doc in collection.documents:
         for passage in doc.passages:
             writer.write(passage)

or

from bioc import toJSON
import jsonlines
with jsonlines.open(filename, 'w') as writer:
    for doc in collection.documents:
         for passage in doc.passages:
             writer.write(toJSON(passage))

Incrementally decoding the BioC Json lines file:

from bioc import BioCJsonIterReader
with open(filename, 'r', encoding='utf8') as fp:
    reader = BioCJsonIterReader(fp, level=bioc.PASSAGE)
    for passage in reader:
        # process passage
        ...

or

from bioc import fromJSON
import jsonlines
with jsonlines.open(filename) as reader:
    for obj in reader:
        passage = fromJSON(obj, level=bioc.PASSAGE)
        ...

Development

Test

$ pytest cov=bioc tests

Publish BioC to PyPI and TestPyPI

First, you need a PyPI user account. You can create an account using the form on the PyPI/TestPyPI website.

Now you’ll create a PyPI/TestPyPI API token so you will be able to securely upload your project.

Go to https://pypi.org/manage/account/#api-tokens and create a new API token; don’t limit its scope to a particular project, since you are creating a new project.

$ python -m build
$ python -m twine upload --repository testpypi dist\*

Developers

• Yifan Peng ([e-mail])

Webpage

The official BioC webpage is available with all up-to-date instructions, code, and corpora in the BioC format, and other research on, based on and related to BioC.

• http://www.ncbi.nlm.nih.gov/CBBresearch/Dogan/BioC/

• http://bioc.sourceforge.net/

Reference

If you use bioc in your research, please cite the following paper:

• Peng Y, Tudor CO, Torii M, Wu CH, Vijay-Shanker K. iSimp in BioC standard format: Enhancing the interoperability of a sentence simplification system. Database (Oxford). 2014, 1-8. bau038.